基于平方根UKF的水下纯方位目标跟踪

为了避免被动跟踪中非线性性带来的计算复杂化及跟踪精度的下降,该文将平方根无迹卡尔曼滤波(SR-UKF)算法应用到水下仅测角目标跟踪.利用协方差平方根代替协方差参加递推运算,解决了标准无迹卡尔曼滤波(UKF)算法中由于计算误差和噪声等因素有可能引起误差协方差矩阵负定而导致滤波结果发散的问题,保证了滤波算法的数值稳定性,提高了跟踪的精度和可靠性.仿真结果表明,SR-UKF非线性滤波算法应用于水下仅测角目标跟踪系统是有效的,而且滤波精度、稳定性和收敛时间明显优于扩展卡尔曼滤波(EKF)和标准UKF算法.

Underwater Bearing-only Target Tracking Based on Square-root UKF

To avoid the computational complexity and the tracking precision decrease from the nonlinear feature in passive tracking, a new square-root unscented Kalman filter(SR-UKF) algorithm is proposed to track underwater targets. The covariance square root matrix is taken in stead of covariance matrix in filter recursion. The filtering divergence problem caused by non-positive error covariance matrix in general unscented Kalman filter(UKF) is solved, and the tracking precision and stability of the algorithm is improved. The simulation results show that the SR-UKF is an effective nonlinear filtering method for underwater bearing-only tracking system, and it performs better than extended Kalman filter(EKF) and general UKF in filtering precision, stability and convergence time.